Dwight I'll offer my own take on your questions. 1. Think of the two Chr like a stone wall with two layers of stones of random sizes. So, yes, if two shared segments line up closely they will more often be from the same Chr. That's *more often*, not *always*. Your best bet is to do everything you can to determine Triangulation. A Triangulated Group is virtually always from one Chr. 2. GEDmatch is a great tool for Triangulation. You know the ID numbers - compare them 1-to-1 - it's that simple! I've checked many, many GEDmatch matches against known relatives on both sides and come up with many small shared segments (in 1-5cM range) with the Match on both sides. I don't put any stock in the *many small segments* concept of matching. After many 1-to-1 looks - I haven't really detected a trend. Even if analysis showed a 30-70 trend, that is not good enough for me to make the assignment (I've gotten too convoluted and mixed up in the past with such assumptions). Stick with Triangulation. Jim - Sent from my iPhone - FaceTime! On Dec 18, 2013, at 12:55 PM, Dwight Holmes <dwightrholmes@gmail.com> wrote: > Tim - > Thanks, as always, for your thoughtful, informative responses. I'd like to > ask a couple of questions for my own benefit, and I'll assume others may > also benefit from this (including Bob, whose thread this is!) > > 1. I know that segment start and stop points are approximations. But does > the fact that the two segments have the exact same reported start and stop > points, cM, and SNPs, add to the likelihood they are the same DNA segment? > 2. As to the question of whether these are on the same or opposite sides of > Ch 3, am I right that Gedmatch cannot answer that question directly? But > that you might see in the data for the four people other evidence (a large > number of smaller shared segments, for example) that might point to them > all being related, or (in the absence of such supporting indications) not? > > Thanks! > > Dwight >